418
ERIC P. CHASSIGNET ET AL.
3.
The North Atlantic prototype ocean prediction system
While HYCOM is a highly sophisticated model, including a large suite
of physical processes and incorporating numerical techniques that are
optimal for dynamically different regions of the ocean, data assimilation is
still essential for ocean prediction a) because many ocean phenomena are
due to flow instabilities and thus are not a deterministic response to
atmospheric forcing, b) because of errors in the atmospheric forcing, and c)
because of ocean model imperfections, including limitations in resolution.
One large body of data is obtained remotely from instruments aboard
satellites. They provide substantial information about the ocean’s space-time
variability at the surface, but they are insufficient by themselves for
specifying the subsurface variability. Another significant body of data is in
the form of vertical profiles from XBTs, CTDs, and profiling floats (e.g.,
ARGO). While these are too sparse to characterize the horizontal variability,
they provide valuable information about the vertical stratification. Even
together, these data sets are insufficient to determine the state of the ocean
completely, so it is necessary to exploit prior knowledge in the form of
statistics determined from past observations as well as our understanding of
ocean dynamics. By combining all of these observations through data
assimilation into an ocean model it is possible to produce a dynamically
consistent depiction of the ocean. It is important that the ocean model
component of the forecast system has skill in predicting the ocean features of
interest. Then the model can act as an efficient dynamical interpolator of the
observations.
Performance of HYCOM in the North and Equatorial Atlantic has been
documented by Chassignet et al. (2003) within the framework of the
Community Modeling Experiment (CME). The near real time 1/12º (~7 km
mid-latitude resolution) HYCOM Atlantic Ocean prediction system
(http://hycom.rsmas.miami.edu/ocean_prediction.html) spans from 28
o S to
70
o N, including the Mediterranean Sea and has been running since July
2002. The vertical resolution consists of 26 hybrid layers, with the top layer
typically at its minimum thickness of 3 m (i.e., in fixed coordinate mode to
provide near surface values). In coastal waters, there are up to 15 sigmalevels, and the coastline is at the 10 m isobath. The northern and southern
boundaries are treated as closed, but are outfitted with 3
o buffer zones in
which temperature, salinity, and pressure are linearly relaxed toward their
seasonally varying climatological values. Three-hourly wind and daily
thermal forcing (interpolated to three hours) are presently provided by the
FNMOC Navy Operational Global Atmospheric Prediction System
(NOGAPS) (Rosmond et al., 2002), available from NAVOCEANO and the
U.S. GODAE data server in Monterey. The HYCOM prediction system uses
ERIC P. CHASSIGNET ET AL.
3.
The North Atlantic prototype ocean prediction system
While HYCOM is a highly sophisticated model, including a large suite
of physical processes and incorporating numerical techniques that are
optimal for dynamically different regions of the ocean, data assimilation is
still essential for ocean prediction a) because many ocean phenomena are
due to flow instabilities and thus are not a deterministic response to
atmospheric forcing, b) because of errors in the atmospheric forcing, and c)
because of ocean model imperfections, including limitations in resolution.
One large body of data is obtained remotely from instruments aboard
satellites. They provide substantial information about the ocean’s space-time
variability at the surface, but they are insufficient by themselves for
specifying the subsurface variability. Another significant body of data is in
the form of vertical profiles from XBTs, CTDs, and profiling floats (e.g.,
ARGO). While these are too sparse to characterize the horizontal variability,
they provide valuable information about the vertical stratification. Even
together, these data sets are insufficient to determine the state of the ocean
completely, so it is necessary to exploit prior knowledge in the form of
statistics determined from past observations as well as our understanding of
ocean dynamics. By combining all of these observations through data
assimilation into an ocean model it is possible to produce a dynamically
consistent depiction of the ocean. It is important that the ocean model
component of the forecast system has skill in predicting the ocean features of
interest. Then the model can act as an efficient dynamical interpolator of the
observations.
Performance of HYCOM in the North and Equatorial Atlantic has been
documented by Chassignet et al. (2003) within the framework of the
Community Modeling Experiment (CME). The near real time 1/12º (~7 km
mid-latitude resolution) HYCOM Atlantic Ocean prediction system
(http://hycom.rsmas.miami.edu/ocean_prediction.html) spans from 28
o S to
70
o N, including the Mediterranean Sea and has been running since July
2002. The vertical resolution consists of 26 hybrid layers, with the top layer
typically at its minimum thickness of 3 m (i.e., in fixed coordinate mode to
provide near surface values). In coastal waters, there are up to 15 sigmalevels, and the coastline is at the 10 m isobath. The northern and southern
boundaries are treated as closed, but are outfitted with 3
o buffer zones in
which temperature, salinity, and pressure are linearly relaxed toward their
seasonally varying climatological values. Three-hourly wind and daily
thermal forcing (interpolated to three hours) are presently provided by the
FNMOC Navy Operational Global Atmospheric Prediction System
(NOGAPS) (Rosmond et al., 2002), available from NAVOCEANO and the
U.S. GODAE data server in Monterey. The HYCOM prediction system uses